Tag
A large-scale survey reveals that U.S. workers overestimate the risk of job loss from generative AI, with fears not reflected in actual labor market data; instead, fear correlates with higher AI usage for core job functions.
A Penn State-led study published in PLOS One finds that over half of U.S. adults report lacking basic statistical knowledge, yet most would rely on statistics more if they understood them better.
This paper develops a staged robustness analysis framework that connects Structural Equation Modelling (SEM), Ordinary Least Squares (OLS), and Double Machine Learning (DML) for survey-based latent-construct research, demonstrated on a FinTech Digital Customer Intimacy survey model. The framework provides a reusable template for researchers to assess stability of findings across different estimation methods.
This paper presents a five-stage framework integrating large language models into survey research, addressing declining response rates, sample bias, and fraudulent completions. Using 2024 Hurricane Milton survey data, the authors propose a theory-informed LLM (A-TLM) that outperforms classical imputation methods in missing-data scenarios and demonstrates manageable hallucination risk through grounded refusal.